DESCRIPTION: (Adapted from applicants abstract) Over 1 million persons will be diagnosed with cancer next year. A disproportionate number will be African-Americans who have greater risk, lower survival, and higher death rates from cancer. Recent studies suggest that these findings may be more strongly related to poverty than to racial predisposition. Yet, the representation of minorities and other underserved populations in clinical trials is very low. This competing continuation proposal will determine if the outcomes of a theory-driven randomized trial of nursing interventions for radiotherapy patients can be replicated when the target population is culturally, racially, educationally, and socioeconomically diverse. The proposal also explores the effects of spouses' participation on patient outcomes. Prior findings with a middle-class, well educated Anglo-American sample indicate that informational and/or behavioral interventions are associated with more knowledge about radiotherapy and its side effects, more self-care actions to manage those side effects, decreased mood disturbance, and increased satisfaction with care. In the study of a demographically heterogeneous sample of head and neck cancer patients (43% minority), similar results are seen when reinforcement contracting, a behavioral strategy that corresponds well to Chrisman's definition of culture- sensitive care, is used with information. In the proposed study, patients will be stratified by treatment site, gender, and race, and randomized to either Information Alone, Information with Contracting, or Standard Care (control). The experimental groups will be further stratified by presence/absence of a spouse willing to participate, with 50% randomly selected to be included with the patient. Patients and the selected spouses will be seen by a clinical nurse specialist three times during their course of therapy; data collection occurs at baseline, at the end of treatment, at first follow-up visit, and at 6 months. Dependent variables include patients' self care ability, self care behavior, functional health, emotional distress, quality of life, general life satisfaction, satisfaction with care and spousal and self care concerns. Predictor and moderator variables include demographics, risk for distress, ethnic affiliation, treatment site, number and severity of side effects, and spouse variables pertaining to spousal and self care concerns, caregiving demands, knowledge, and quality of life. The study will also examine the extent to which the psychometric properties of a pretreatment nursing assessment and its ability to predict later emotional distress are retained in this diverse populations.